To appropriate the famous Martin Mull quote, writing about visualization is like dancing about architecture. Why are we using written words, like this blog post, to talk about visualization instead of moving images, like in a video?
It gets weirder, because the impetus for writing this comes from a discussion on Clubhouse, the audio-only social media/hangout place that’s all the rage right now. Yes, we were discussing visualization without any access to visuals. I’m just as confused as you are.
But seriously, visualization is obviously a visual medium, and yet there isn’t much visualization content that is video. And by video, I mean YouTube. The comment that started this post was about how many people now search for information on YouTube, and even if they use a more general search engine expect to find videos. I know I’ve been doing that more and more lately. And I’m clearly not alone, check out some staggering stats: two billion logged-in monthly users watching a billion hours of video every day. YouTube is the second-most popular website out there (only beaten by Google).
Visualization Visuals
Video being a visual medium, well-designed and executed visuals are clearly a way to get an audience. 3Blue1Brown is an outstanding example in my opinion, using extremely well-designed and recognizable animations to explain complicated mathematical ideas. A few other random examples that come to mind (and that I happen to know) are Practical Engineering and CGP Grey.
None of them use data visualization, but they all use visuals for part or all of their videos to explain and illustrate ideas and concepts. Visuals work well in video for obvious reasons, but I think it’s helpful to see some of these examples to appreciate just how well they can be designed.
Where does data visualization fit in? Just like explaining engineering concepts, mathematical ideas, or political systems can be done using illustrations, many other things could be argued or explained using data. And of course, visualization itself can often use some visual help to help people understand how to read it, what to compare to, etc.
Niches
But, you say, visualization is a niche! How can I expect to get traction on something that isn’t makeup tutorials or some kind of gadget reviewing? Fair point.
Know what else is a niche? Music theory. And yet, Adam Neely has close to 1.5 million subscribers, 12tone has over 400,000, David Bennett has 370,000, and David Bruce over 180,000.
Yes, music theory. Now tell me again about visualization being a niche? There are much smaller niches on YouTube with massive numbers of people watching. Even a small niche translates into a large number of people when the total possible audience is in the billions.
Video Is Too Much Effort
Yes, it is a lot of effort to create videos. Mine take many hours, and while you can certainly just mumble into your camera for 20 minutes and upload that, you won’t get much traction.
I do think that it’s possible to make good videos with less effort, though. When I first thought about making videos several years ago, I initially thought I’d draw my visuals by hand, xkcd-style. Alas, my drawing is not exactly of the quality I would be happy to use in a video. But other people surely can do better and could do great things with just paper and a pen (or an iPad and a drawing program).
There are some very successful channels on YouTube that use hand-drawn visuals. Two that come to mind are 12tone and minutephysics. 12tone is really just hand-drawn, though certainly with a lot of thought put into his scripts. And minutephysics animates and manipulates some of the drawings, but most of it is just based on (relatively) simple but well thought-out drawings.
Quality has many aspects and can be achieved in different ways. There are the absurdly high production values of Captain Disillusion, but there are many others who make good videos that people enjoy because of their content even when the visual quality isn’t that great.
Tutorials, Explanations, Research, Etc.
I think there is a lot of room on YouTube for visualization content. Some of it basically presents itself: how do I do X in software Y? Tutorials are probably the most obvious things people look for on YouTube, and they’re also the easiest to do. They do require planning, scripting, and pacing though, or they get incredibly boring and tedious. I’m a huge fan of Ripple Training for Final Cut Pro tutorials, and their 60-second videos in particular. There’s an art to making tutorials that aren’t just informative, but also enjoyable.
But there is more to visualization than the pure mechanics. I’m obviously trying to do more conceptual things on my channel, and there’s a lot more to say about the thinking behind data analysis, data preparation, statistics, visualization, etc.
And then there’s research. A lot of people are really interested in learning more about what is going on in the research world, but their access is quite limited or they don’t know where to go. Videos accompanying papers are more common now, but they often don’t stand on their own, aren’t paced or scripted very well, are filled with jargon, etc.
But the bones, as it were, are there. There’s clearly a lot of potential in bringing more visualization to people through video. And I think it would be a great exercise for students to think about structuring how they present their work, not just for video but also in writing.
Isn’t it weird that there are more successful podcasts about visualization than video channels? Forget Clubhouse, YouTube is the natural habitat for visualization.
13 responses to “The Dearth of Videos about Visualization”
We do not write about visualization in words, we write about visualization in pictures. Words are just an addition. In my opinion the most successful format for complex data visualization is scrollytelling. This allows you to move forward, backward and delve into each and every part. The problem with video is the difficulty in moving back and forth and managing viewing speed. I can read the articles at my own pace, I must watch the video at the pace of the creator.
I’m working on it! :) https://youtube.com/c/AnnKEmery
I mostly create videos for online courses though. I have 100 hours of training that 4,400 people have joined so far. https://depictdatastudio.teachable.com/
Hi Robert. Thanks for a great, thought-provoking post.
It’s very interesting for me because I decided recently to write about data vis, presenting and similar topics. I questioned whether blogging was the right medium because I kept stumbling upon podcasts, or practitioners using YouTube.
I think there’s a place for writing about data vis. I’ve learned a huge amount from reading your posts, books by Alberto Cairo and others, Lisa C Rost’s blog and so on. My early posts will be ‘thought pieces’, describing concepts and ways to think about data vis. There will be some images but enough to justify making a video? I don’t know. It would mostly be me talking to camera (or static images).
YouTube is great for walk-throughs that teach a skill so you can follow the steps. I may do some. It’s worth noting that content like that can get bundled into courses delivered somewhere else, e.g. SkillShare. However they’re shared I’m with you; sometimes they’re the best way.
Podcasting and ClubHouse are the intriguing choices. Tackling data vis with no visuals seems like an odd choice. But they can be strangely engaging. Cole Nussbaumer Knaflic’s podcast is great in how she brings others into the discussion and makes any visuals available as a complement to the show. I listen to them while running (not so easy to do with videos).
There’s lots of content around now so it’s interesting to see how it’s being done.
I am try to make channel about data journalism. its very hard https://www.youtube.com/channel/UCQ3cyKr_SnAb5Jsfwa3VCjw
So why is this post not a video, then? :-)
I’d create a video to accompany every blog post that I write but, as you point out, it’s hideously time-consuming to make something that’s actually worth watching (it takes me at least three to five full days of work for a three- to fifteen-minute video). While video is almost always more effective (and not just for data viz…), sometimes it adds a lot more value than others, depending on the specific topic. My rule is that, if a video will make a big difference for the topic at hand, I’ll create one, but if it’s a topic for which a video wouldn’t add as much value, I’ll leave it as a textual blog post to be able to devote the saved time to other projects.
Aha, but I did not say everything has to be a video ;) Quite the opposite actually, I think it’s important to choose the right medium for the topic and material you have. Some things work better as written text, like my posting here. There aren’t a lot of visuals to illustrate my points. But in other cases, like when discussing visualizations, there are (or should be)
Good point, although the question then becomes when text is more effective than even just a talking head saying basically the same thing with no visuals (if none are needed, although even just some slides with a few bullets are often more effective than just a talking head with no visuals). The main advantages I see with text are that it’s usually faster to consume and it offers better “random access” to refer back to something that was encountered earlier on in the piece. Apart from that, it’s usually less engaging and requires more cognitive effort to consume than video because the audience has to “decode” text into spoken language in order to understand it (i.e., why kids often move their lips when they read). I suspect that this is probably why podcasts, YouTube, audiobooks, etc.are so popular now, and print media tends to be becoming less popular (for better or for worse).
I think it comes down to editing more than talking/video vs. text. Text is almost always at least read over once, but a lot of really pointless videos are just unstructured talking. I can’t stand those, and I also don’t listen to podcasts that aren’t reasonably well structured. They just feel like a waste of time.
So I think the difference isn’t just that video needs to have a reason, like good visuals (not just slides with talking points that add no value), but that it needs to be scripted and edited. Text is sort of edited by default, and video needs to be considered and edited too to be worth it.
Ah, but the question has now changed from “When does text tend to be more effective than video?” to “Does text tend to be more carefully crafted than video?” :-) That’s a different question to which I’d have a different answer (which would be similar to yours).
Yes, a poorly structured video is going to be less effective than a well-structured piece of text. The more interesting comparison is between a well-crafted video and well-written piece of text…
True, but I think the two questions are quite strongly intertwined. But there’s also the “is it worth it?” question, and there I just think that a short piece of text (think a tweet or short blog post) is going to be much better bang for buck (storage, effort, clicks, time to consume, annoyance if it wasn’t worth it, etc.) than video in general.
Oh, and to your other point: if you’re comparing high-quality video and writing, video is still probably quite a bit more effort. But if it’s done well and you’re going to show a lot of visuals anyway, the video will be worth it.
I’m going to politely disagree with the terms “scripting” and “editing.”
By “scripting,” I think you mean that vides should be well-planned. Just turning on the recorder and word-vomiting is a big waste of our audience’s time. But…. most of us will sound more natural *without* a script vs. with a script. I like to plan in bullet-point format, but then just speak naturally off those bullet points, but not from a word-for-word script.
By “editing,” I think you mean that videos should look and sound professional. But I’d argue that you don’t need any editing skills at all, and you don’t need to hire someone else to edit either. You can make outstanding one-take videos.
Here’s a recent example, which was planned but NOT scripted, and professional (I think? ha) but with zero editing: https://www.youtube.com/watch?v=PCgFntGjglo&feature=emb_title
I certainly agree that scripting doesn’t have to mean word for word. I tend to end up writing full sentences when I’m really thinking through something, but bullet points can totally work. But it’s scripting either way as opposed to the word-vomiting you mention ;)
I do think editing is important unless you’re really good at sticking to your script and don’t go off on side rants like *some of us* are prone to. Editing is actually pretty easy if your recording is decent, I don’t think you should have to pay somebody else to do it.
The video you link to is clearly well scripted (in the bullet-point sense), you have prepared slides, and it seems like you’ve given that talk before. So maybe rehearsing is a possible alternative to scripting too. And it depends on your style and delivery of course, some people can pull it off, others not without extensive editing.